ABSTRACT: Prenatal alcohol exposure (PAE) can lead to a variety of cognitive, behavioral, and health deficits
falling under the umbrella of fetal alcohol spectrum disorder (FASD). FASD is estimated to affect 2-7% of
children in the USA and 23% worldwide. Importantly, people with FASD have much higher rates of mental
health problems than the general population. In particular, up to 50% of people with FASD suffer from
depression, making it one of the most prevalent mental illnesses linked to PAE. Although the biological
mechanisms underlying this increased vulnerability remain unknown, DNA methylation (DNAm) – a type of
epigenetic modification – has emerged as a prime candidate to explain the long-term effects of PAE and its
links to depression. However, most human studies of PAE and its effects on DNAm and depression are cross-
sectional, and thus, have not investigated if the timing of PAE influences these relationships. Here, we propose
to determine the extent to which the timing of PAE influences DNAm and depressive symptoms in childhood
and adolescence, as well as assess the role of DNAm in mediating the link between PAE and increased
depression risk. We will analyze data from the Avon Longitudinal Study of Parents and Children (ALSPAC), a
longitudinal birth cohort that collected repeated, prospective measures of PAE during pregnancy, DNAm and
genetic data, and measures of depression collected almost yearly from age 4 to 16.5. We will replicate findings
in Generation R (GenR), a longitudinal birth cohort with similar metrics to ALSPAC from children followed for
17 years. In Aim 1, we will assess the extent to which the timing of PAE influences parent-reported, child
depressive symptom trajectories from age 4 to 16.5. Trajectories will be characterized using growth mixture
modeling with structured residuals, a method we previously used to identify six classes of depression
trajectories in ALSPAC. Causal relationships will be tested through Mendelian Randomization and negative
control analyses (i.e., partner drinking in pregnancy). In Aim 2, we will identify the DNAm patterns at birth that
are influenced by the timing of PAE and determine the extent to which these DNAm profiles mediate, or
partially explain, the relationship between PAE and depressive symptom trajectories using statistical mediation
methods. Both aims leverage a two-stage structured life course modeling approach previously used by our
team to identify age periods when early-life exposures have greater effects on DNAm and depression. In sum,
this study will identify: (1) periods when PAE has larger effects on depressive symptoms and DNAm; (2)
specific patterns of depression in childhood and adolescence driven by PAE; and (3) epigenetic alterations that
link PAE to depression. These findings will highlight developmental windows and biological mechanisms that
could be targeted in interventions that reduce depression risk among people with FASD and maximize their
well-being. The proposed research will also generate preliminary evidence towards future grant applications
focused on socio-biological factors that link PAE to mental health across the life course.
